Monday, February 21, 2011

More weird corrections

Again taking the final temperature of several towns and subtracting from it, the raw data, we get more of the strange pulsating corrections.







These total correction curves clearly show that the climatologists think that the raw data is as many as 3-5 degrees in error, which is why they correct it that much, yet they tell us that the observations are good enough to measure to the accuracy of a tenth of a degree. I always learned that if you have to make a degree correction, you can't claim that you have tenth of a degree precision.

For Nonymous, please explain why there is a jitter in the error, meaning one must believe that the error is sinusoidal. What causes this sinusoid?

5 comments:

  1. Well, since you were so kind as to call me an unthinking nuisance in the previous posting I'll assume that anything I say or do is without merit. So you can take this with a grain of salt, but when I look through the Karl paper from 1986 (http://journals.ametsoc.org/doi/pdf/10.1175/1520-0450%281986%29025%3C0145%3AAMTETT%3E2.0.CO%3B2)

    It looks to me in figure 1 that the TOBS adjustment actually changes with the month as one might expect. The Metadata for the Albany, TX station seems to indicate a 3:00AM measurement temp (?) so the adjustment between 3 and midnight, one would think, would change with the seasons. If you look at Figure 1 in the Karl paper it appears that this is what is happening. Which may very well explain an adjustment up and down throughout the year.

    This seems to explain the 12-month cyclicity of the TOBS correction, what is your take on this?

    ReplyDelete
  2. Your grammar is appalling.

    ReplyDelete
  3. To anonymous, yes, my grammar is appalling. But content is what is important.

    To Nonymous (love how these guys don't want to be identified). The problem with that explanation is that the MMTS system automatically records the entire day, so there is no Tobs correction required because they record hourly. There is no Tobs bias in the newer MMTS. Yet the high frequency periodicity continues. Thus, the Tobs is not the explanation for this particular periodicity. It can't be.

    And I hope my grammar meets the approval of Anonymous, who seems not to be able to address substance.

    ReplyDelete
  4. Very interesting point. But when you look at the USHCN description of how the data itself is treated it states: "The TOB debiased data are input into the MMTS program and is the second adjustment. The MMTS program debiases the data obtained from stations with MMTS sensors."

    I am hoping to see if I can get ahold of a copy of Quayle 1991 which may give more insight into this adjustment.

    From reading Karl 1986 you can see the inclusion of the month before and after a given data point. I believe the TOBS bias adjustment includes more than merely the diurnal change but also how the measurement is temporally spaced. Since these go into time-based measurements for the trends the TOBS may still be applied.

    If I understand an MMTS system the older (non-thermistor systems) are actually glass thermometers that physically recorded the min or max for some given block of time and had to be reset manually. So any "offset" from a midnight based measurement, it would stand to reason, could still be in need of an adjustment for TOBS.

    ReplyDelete
  5. I was thinking that "Anonymous"'s comment was more aimed at MY grammar, which in my comment was pretty bad.

    ReplyDelete